Characterization of phospho-olivines as materials for Li-ion cell cathodes
Combustion-synthesized sodium manganese (cobalt) oxides as cathodes for sodium ion batteries
Transcript of Combustion-synthesized sodium manganese (cobalt) oxides as cathodes for sodium ion batteries
[1]
Combustion Synthesized Sodium Manganese (Cobalt) Oxides
as Cathodes for Sodium Ion Batteries
Nicolas Bucher,# Steffen Hartung,
# Irina Gocheva, Yan L. Cheah, Madhavi Srinivasan,
*
Harry E. Hoster*
Abstract
We report on the electrochemical properties of layered manganese oxides, with and without
cobalt substituents, as cathodes in sodium ion batteries. We fabricated sub-µm sized particles
of Na0.7MnO2+z and Na0.7Co0.11Mn0.89O2+z via combustion synthesis. X-ray diffraction
revealed the same layered hexagonal P2-type bronze structure with high crystallinity for both
materials. Potentiostatic and galvanostatic charge/discharge cycles in the range 1.5 V…3.8 V
vs. Na Na+ were performed to identify potential dependent phase transitions, capacity, and
capacity retention. After charging to 3.8 V, both materials had an initial discharge capacity of
138 mA h g-1
at a rate of 0.3 C. For the 20th
cycle those values reduced to 75 mA h g-1
and
92 mA h g-1
for Co-free and Co-doped samples, respectively. Our findings indicate that
earlier works probably underestimated the potential of (doped) P2-type Na0.7MnO2+z as
cathode material for sodium ion batteries in terms of capacity and cycle stability. Apart from
doping a simple optimization parameter seems to be the particle size of the active material.
N. Bucher, S. Hartung, I. Gocheva, M. Srinivasan *, H.E. Hoster *
TUM CREATE
Singapore 138602, Singapore
N. Bucher, S. Hartung, H.E. Hoster *
Technische Universität München
85748 Garching, Germany
Y. L. Cheah, M. Srinivasan *
School of Materials Science and Engineering, Nanyang Technological University
Singapore 639798, Singapore
# Both authors contributed equally to this manuscript
[2]
1. Introduction
Growing interest in the integration of sustainable energy sources into the electricity grid has
spurred immense interest in the field of energy storage applications in recent years. The pre-
dominant commercial battery system is the lithium ion battery (LIB), which has become the
standard energy source for portable electronic devices and currently gains increasing
importance for the electric propulsion of vehicles [1–6]. There are several approaches to
remedy the inherent problems of LIBs, like battery cost and safety. One approach that has
attracted wide attention is aqueous lithium ion battery systems. Only recently, several water-
based batteries with high capacities of approximately 120 mA h g -1
and excellent cycle
efficiencies using LiMn2O4 as cathode material have been developed [7, 8][9]. Another
promising alternative to conventional LIBs is the sodium ion battery (NIB) [10–13] owing to
worldwide abundance of Na metal and lower cost of Na2CO3 (~0.40 € kg-1
) than that of
Li2CO3 (~4 € kg-1
) [14]. Even more important are potential cost savings for the cathode
material. Such cost arguments are even more relevant for larger scale rechargeable batteries
as they are currently discussed for the stabilization of electric grids. Though both Li and Na
are alkali metals, the best electrode materials for LIBs are not necessarily the best for NIBs.
In particular, NIBs offer the potential to avoid the expensive cobalt oxide based cathodes that
are common in LIBs [15]. Two main challenges related to NIBs are (i) the generally lower
voltage and (ii) the larger volume change due to (de-)intercalation of Na+ as compared to Li
+
ions. The latter implies additional strain on the crystal structure during cycling, with negative
impact on the cycle life. As Na based chemistry is less well explored than Li based chemistry
in the field of energy storage, there is a good chance that known structures that do not work
in LIBs are found to work in NIBs.
Layered manganese oxides in their manifold structures [16, 17] are used and discussed as
cathode materials in lithium ion [18, 19] and sodium ion [20, 21] batteries since long. For Li
ion batteries, LiMnO2 with its high theoretical capacity of 285 mA h g-1
is an attractive
alternative to the common LiCoO2, also due to the much lower price of Mn as compared to
Co. An inherent problem of LixMnO2, however, is its instability against a transformation
from a layered into a spinel structure in charged state, i.e., at low Li content x [22], which
goes along with capacity fading. Ma et al. recently highlighted that for NaxMnO2 such a
degradation to a spinel structure is thermodynamically and kinetically disfavoured as
compared to LixMnO2 [23], i.e., NIBs could be fabricated based on this rather cheap and non-
toxic class of cathode materials. Nevertheless, cycling stability remains a central challenge
for Na ion batteries [14, 21] and a broad community is tackling this challenge in several ways
[3]
[24–27]. For layered manganese oxides as cathodes in LIBs, significant improvements of the
cycle life were achieved by partial substitution of Mn by other transition metals
(“doping”)[19]. Interestingly, those cathode materials were prepared via the sodium
containing precursors Na0.7Mn0.89M0.11O2+z (M=Fe, Co, Ni, Cu, Zn, Al; z = 0.05 accounts for
cation vacancies in the MO2 layers [16]), but their electrochemical properties were only
tested after an ion exchange from Na to Li.
In this paper, we will show that the powders generated by the combustion synthesis method
[28] used by Doeff et al. in the cited work [19] are not only useful precursors for the
fabrication of LIB cathodes but are also well-performing cathodes for NIBs. As test cases, we
will show first results for Na0.7MnO2.05 and Na0.7Mn0.89Co0.11O2.05, both with the same P2-
type layered structure [17]. After a description of synthesis and analysis methods and
procedures we will first present the crystal structure and morphology of the samples as
analyzed by X-ray diffraction (XRD) and Field Emission Scanning Electron Microscopy
(FESEM), respectively. We will then evaluate potentiostatic and galvanostatic charge-
discharge curves of both materials with respect to their initial Na content, their potential
depending phases, their cycle efficiency, and their cycling stability. The results will be
discussed in comparison to previous findings for P2-Na0.7MnO2 [20], P2-Na0.6MnO2[21],
monoclinic NaMnO2 [23], P2-NaxCoO2 [29], and P2-Na2/3Co2/3Mn1/3O2 [30].
2. Experimental
2.1 Synthesis
Na0.7MnO2+z was synthesized using the combustion method as described by Doeff et. al. [19,
28, 31, 32]. Sodium nitrate (Sigma Aldrich, ≥ 99%) and Manganese acetate (Alfa Aesar,
anhydrous, 98%) were mixed in Na:Mn mole ratio of 7:10 before being dissolved in
deionized (DI) water, followed by the addition of concentrated nitric acid (≥ 69%,
Honeywell) in excess and 1.5 g of gelatin. The solution was heated to 250 °C until
spontaneous combustion occurred. Oxidation of the amino groups of the gelatin not only
releases the required heat but also yields substantial amounts of gaseous products (N2, CO2,
H2O) that facilitate the reaction and ensure small particle sizes. The product is a dark
brownish powder with cotton-wool-like texture. For better crystallization, the powder was
annealed in air at 800 °C for 4 h. For the cobalt-doped cathode, cobalt nitrate (Sigma Aldrich,
≥ 98%) was added to the precursors in the mole ratio of Na:Co:Mn 7:1.1:8.9. The subsequent
synthesis steps were carried out as described above, obtaining Na0.7Co0.11Mn0.89O2+z
[4]
(according to elemental analysis). In this paper, we will refer to the Co-free and Co-doped
sample as NMO and NCO, respectively.
2.2 Characterization
The crystal structures of the synthesis products were analyzed by powder XRD using a
Bruker X-ray diffractometer with Cu-Kα radiation. 2θ was varied within a range of 10o < 2θ <
80o, using a step size of 0.017
o and 1.7 s per step. The obtained XRD patterns were analyzed
by Rietveld phase analysis using a Lebail approach within the Topas version 3 software,
employing the fundamental parameters approach [33–35]. The powder morphology was
analyzed by FESEM (JEOL JSM-7600F), including elemental analysis. Brunauer-Emmet-
Teller (BET) surface area measurements were carried out with a Nova 3200e surface area and
porosity analyzer.
2.3 Electrochemical measurements
The composite cathodes were prepared by mixing active material (NMO or NCO), acetylene
black (Alfa Aesar, > 99%), and polyvinylidene fluoride (PVdF, Arkema, Kynar HSV 900)
binder in the weight ratio of 60:25:15, with N-Methyl-2-pyrrolidone (NMP, Sigma Aldrich,
≥ 99%), to form a slurry. The well-mixed, homogenous mixture was coated on an Al foil
(16 m in thickness) using a doctor blade, and dried at 80 °C in air to remove the solvent.
The dried coating was then roll-pressed between twin rollers to improve adherence to the Al
foil, and then punched into 16 mm diameter pieces. After drying the electrodes for 4 h under
vacuum at 110 °C, the electrodes were assembled in half-cell configuration in 2016 coin
cells, using 16 mm circular sodium metal pieces as the anode, separated by glass fiber
(Whatman) swollen with sodium-ion conducting electrolyte. That electrolyte consisted of 1M
NaClO4 (Sigma Aldrich, ≥ 98%) in propylene carbonate (PC, Sigma Aldrich, ≥ 99.7%).
Electrochemical studies were carried out via cyclic voltammetry using a BioLogic
potentiostat while galvanostatic testing were performed using a battery tester system
(Neware). All capacities are reported in mA h g-1
, related to the mass of the active material.
For Na0.7MnO2 with a molar mass of 102.41 g mol-1
the extraction/insertion of 1 Na per
formula unit is equivalent to 262 mA h g-1
.
3. Results and Discussion
The crystal structure of P2-Na0.7MnO2 (space group P63mmc, see ref. [17] for the
nomenclature) is depicted in Figure 1. It consists of MnO2 slabs made of edge-sharing MnO6
octahedra. The sodium ions can occupy two different triangular prismatic sites in the plane
[5]
between these MnO2-slabs. Following a nomenclature introduced by Carlier et al. [30] for a
similar system, these are marked as Naf and Nae in Figure 1. For Naf, the triangles of the
prisms are also triangles of the octahedra above and below, whereas the triangles of the
prisms around Nae are formed by edges of three of the octahedral, respectively. Energetically,
the two sites are not equivalent due to their differing number and distance of the Mn3+/4+
ions
surrounding the respective Na+. The lower right corner of Fig.1 exemplarily depicts a
substituting CoO6 octahedron as it appears in NCO. Based on the results of a neutron
diffraction analysis published for P2-Na2/3Co1/3Mn2/3O2 [36], we assume that the lateral
distribution of CoO6 and MnO6 octahedra within the slabs of our Na0.7Co0.11Mn0.89O2+z does
not obey any long-range order. As mentioned above, the MO2 layers have about 1% cation
vacancies [16]. Such a vacancy is illustrated next to the MnO6 octahedron in Fig. 1.
All peaks in the XRD patterns in Figure 2 can be indexed to the layered P2 structure of
Fig. 1. For the lattice vectors a, b, c, a Rietveld analysis yields the lattice parameters a = b
=2.872(2) Å and c = 11.167(2) Å for NMO, and a = b = 2.848(1) Å and c = 11.149(2) Å for
NCO. FESEM images as those in Fig. 3 revealed that both NCO and NMO consist of flat
hexagonal particles after synthesis. This fits to the layered crystal structure detected in XRD.
The lateral dimensions (parallel to lattice vectors a and b) of the flakes in Fig. 3 are in the
range between 300 … 600 nm and their thickness (direction of lattice vector c) lies between
100 nm … 200 nm. Energy-dispersive X-ray spectroscopy (EDX) confirmed Na, Mn, O, and
(for NCO) Co as main constituents of the electrodes. Quantitative analysis revealed a ratio of
cobalt to manganese of 1:6.7, resulting in a stoichiometry of Na0.7Co0.11Mn0.89O2+z for NCO.
Apart from the microstructure the two samples also share the same specific surface area of
~50 m2 g
-1 as measured by BET. In total, the two samples are essentially equal in crystal
structure and morphology and thus provide an ideal model system to separately study the
effect of the Co substituents in NCO on the electrochemical properties.
After assembly the coin cells with either material had an open circuit voltage (OCV) of
2.65 V. This indicates that both cathodes have a comparable Na+ content after synthesis. The
voltage is close to the 2.8 V reported by Caballero for P2-Na0.6MnO2 obtained by a sol-gel
method [21]. We rationalize our slightly lower OCV as a consequence of a higher Na+
content of 0.7 as opposed to 0.6 in ref. [21]. As will be shown more below, both of our tested
materials take up a charge of 56 mA h g-1
when charged up to 3.8 V. This would be
equivalent to an extrusion of 0.21 Na+ per formula unit, i.e., forming Na0.49MnO2+z and
Na0.49Co0.11Mn0.89O2+z, respectively.
[6]
Figure 4 shows cyclic voltammograms (CVs) of NMO and NCO at a scan rate of 0.3
mV s-1
in the range between 1.5 – 3.8 V vs. Na | Na+. We have plotted the second cycle, but
the first and the later cycles are similar. Both CVs are dominated by a couple of redox peaks
between 2 V and 2.5 V. Both peak positions and the apparent equilibrium potential are
slightly more positive for NCO as compared to NMO. A main difference between the two
voltammograms is the presence of manifold oxidation and reduction peaks for NMO in Fig.
4a as opposed to the rather smooth voltammogram of NCO in Fig. 4b. Similar features were
previously shown in plots of the differential specific capacity of P2-Na0.6MnO2 [21], and also
for monoclinic NaMnO2 in potentiostatic intermittent titration (PITT) patterns [23]. These
features can be attributed to phase transitions similar to those intensely studied for P2-
NaxCoO2 [29, 30, 37, 38]. These transitions do not concern the layered P2 backbone structure
itself but the ordering of Na+ and vacancies between the slabs, specifically their distribution
among the two different triangular prismatic sites (see Fig. 1). Apart from the main redox
couple, the most pronounced features in the CV in Fig. 4a are the two oxidation peaks A and
B at 3.5 V and 3.6 V, respectively. These features seem to be a common fingerprint for
potentiostatic and galvanostatic charging of sodium layered oxides: they can be found, e.g., in
the charging curves for P2-Na0.6MnO2 [21], monoclinic NaMnO2 [23], P2-NaxCoO2 [29, 37],
or P2-NaxCo2/3Mn1/3O2 [30], or P2-Nax[Fe1/2Mn1/2]O2 [39]. For P2-NaxCoO2 [29, 37] and P2-
NaxCo2/3Mn1/3O2 [30] this feature was already shown to be related to the transition into/out of
the very stable phase for an Na+ content of 0.5. For P2-Na0.5CoO2, experiments [40, 41] and
calculations [37, 38, 42] revealed an equal occupancy of Nae and Naf sites. Though the
Na+/vacancy patterns for the corresponding Manganese lamellar oxides are not yet known,
similar ordered and stable structures must exist. As a last point related to the CV profiles in
Figs. 4a and 4b we would like to underline the similarity to the charging curves of P2-
NaxCoO2 [29] and P2-NaxCo2/3Mn1/3O2 [30], where partial substitution of Co by Mn “erased”
most of the features apart from the ones around 3.5 V. Our results show that a continuous
increase of the Mn content towards P2-NaxMnO2+z eventually “revives” those features again.
This can be explained as follows: The interaction of Na+ ions with the chemically corrugated
landscape of the doped oxides predominates over the interactions between the Na+ ions. The
latter is mainly responsible for the ordered patterns observed at specific Na+ concentrations.
Even for plain P2-NaxMnO2+z the Na+ ions can be expected to be in a chemically less
homogeneous landscape as compared to P2-NaxCoO2. The cation vacancies in the MnO2
layers (see Fig. 1) [16] act as effective negative charges that trap intercalated Na+ ions [43]
and thus disturb their arrangement into long-range ordered phases. This would explain why
[7]
the voltammetric features observed in Fig. 4a are still relatively weak despite the good
crystallinity of the cathode material.
Fig. 5 presents galvanostatic charge/discharge cycles of NMO and NCO at 0.3 C (50
mA g-1
) between 1.5 – 3.8 V vs. Na | Na+. For NMO, two potential plateaus at 3.5 V and 3.6
V are visible for all cycles. In analogy to the respective features in the CVs in Fig. 4, these
have been marked as A and B. For the second and all further cycles, the charge curves of
NMO exhibit a long plateau at ~2.3 V as dominating features. In agreement to the CVs, the
corresponding discharge plateau is at ~2.1 V. As to be expected from the much broader peaks
in the CVs of NCO, the galvanostatic charge and discharge curves of that sample in Fig. 5b
do not exhibit any pronounced plateaus. NMO and NCO both exhibit a discharge capacity of
138 mA h g-1
after initial charging to 3.8 V. This is equivalent to the intercalation of 0.53 Na+
per formula unit. The capacity is close to the values reported for P2-Na0.6MnO2 (140 mA h
g-1
[21]), P2-NaxCo2/3Mn1/3O2 and P2-NaxCoO2 (both 120 mA h g-1
at C/100 [30]). Only
monoclinic NaMnO2, where up to 0.8 Na+ per formula unit can be reversibly (de-)intercalated
has a significantly higher capacity of 185 mA h g-1
(measured for C/10) [23].
After 20 cycles, NMO and NCO have a remaining capacity of 75 mA h g-1
and 92 mA
h g-1
, respectively. As illustrated in Fig. 6a, the decay in capacity becomes less steep with
increasing cycle number for both cathode materials. This behavior is different from that
reported by Morales et al. for P2-Na0.6MnO2 [21], where the degradation seems to become
faster with increasing number of cycles. Since their crystal structure should be the same as
ours and the voltage range used in their cycles was even narrower, the most likely
explanation for the better stability of our NMO is the electrode structure, specifically, the
particle size. Whereas the flakes generated by our combustion method are all < 1 µm (cf. Fig.
3), the particles in the cited work are > 2 µm. Smaller particles are less sensitive to
mechanical deformation due to Na+ (de-)intercalation. This would explain why we observe a
better stability and it calls for a systematic optimization of the particle size in future studies.
However, as Qu et al. [44] recently reported for α-NaMnO2 used in asymmetric aqueous
supercapacitors, high cycle efficiencies can also be achieved for larger particles. This might
be a hint that part of the fading observed in our work is due to an interaction between the
organic electrolyte and the active material.
An interesting quantity to consider in the context of capacity retention is the cycle
efficiency, i.e., the discharge capacity divided by the preceding charge capacity. Fig. 6b
shows that the cycle efficiencies for both cathode materials are slowly increasing from 78%
in the second cycle to at most 95% in the twentieth cycle. There is no apparent difference
[8]
between NMO and NCO with respect to these values. Similar behaviour, yet not quantified
over many cycles, was reported for P2-Na0.6MnO2 [21], monoclinic NaMnO2 [23], P2-
NaxCo2/3Mn1/3O2 [30], and P2-NaxCoO2 [29]. The main contribution to the low coulomb
efficiency comes from side reactions at higher potentials, e.g., electrolyte decomposition [13,
23]. Our results do not show any effect of Co-doping on these parasitic reactions. Another
possible reason for capacity fading could be the disproportionation of Mn3+
into Mn4+
and
dissolved Mn2+
, which is a general problem of Mn containing cathodes. This hypothesis is
substantiated by literature [1, 6] and is a key issue to be addressed in future works.
The better capacity retention for the doped sample fits to corresponding observations
for the stability of substituted layered manganese oxides in lithium ion batteries [32]. It has to
be pointed out, however, that the ion exchange of Na+ towards Li
+ goes along with a
transition of a P2 to an O2 structure [32], so that the structural boundary conditions cannot be
directly compared. Moreover, the capacity fading of LIBs involving LixMnO2 cathodes is
mainly due to a gradual transition into an energetically favourable spinel crystal structure [45,
46]. This transition does not occur in NIBs [23]. The stabilizing effect of Co could involve
more than one factor, similar to what was observed for spinel LiCoyMn2-yO4 as compared to
LiMn2O4 as cathodes in LIBs where Co was shown to improve the general chemical stability
and the cation conductivity of the electrode [6].
4. Conclusions
The gelatin assisted combustion method is ideal for the synthesis of P2-type layered
sodium manganese oxides as well-defined model cathodes for rechargeable Na+ ion batteries.
These cathodes can be systematically varied by partial substitution of Mn by Co, Fe, Ni, Cu,
Zn, Al. The resulting powders consist of sub-µm flakes with a P63mmc crystal structure.
Being charged to 3.8 V vs. Na | Na+, the cathodes based on those particles have a discharge
capacity of 140 mA h g-1
, equivalent to the extrusion and re-insertion of ~0.5 Na. Due to the
small size of the flakes, these capacities were attainable at a rate of 0.3 C. We have
exemplarily shown that substitution by Co does not alter the capacity itself but has a
beneficial effect on the cycling stability. Specifically, the capacity of the 20th
discharge cycle
(with 0.3 C) increases from 75 mA h g-1
for Na0.7MnO2+z to 92 mA h g-1
for
Na0.7Co0.11Mn0.89O2+z. An effect of Co doping towards improved cycle efficiency via
suppression of parasitic side reactions, however, can be excluded. Cyclic voltammograms
indicate that the presence of Co suppresses the formation of ordered Na+ / vacancy phases
which is likely to have an influence on the Na+ transport within the cathode. Further studies
[9]
will reveal in how far this behaviour directly contributes to the observed cycle life
improvement. Moreover, a systematic study on the optimum particle size for plain and doped
NaxMnO2 is on its way.
Acknowledgements
The authors would like to thank Miss Han-Yi Chen, Miss Yin Ting Teng and Mr. Jan Geder
for their valuable help in conducting SEM, XRD, and BET measurements. This work was
financially supported by the Singapore National Research Foundation under its Campus for
Research Excellence and Technological Enterprise (CREATE) programme.
[10]
References
1. Wu Y, Wan C, Jiang C, Fang S (2002) Introduction, Principles and Advances of
Lithium Secondary Batteries. Tsinghua University Press, Beijing
2. Scrosati B, Garche J (2010) J Power Sources 195:2419–2430
3. Liu C, Li F, Ma L, Cheng H (2010) Adv Mater 22:E28–62
4. Ceder G, Hautier G, Jain A, Ong SP (2012) MRS Bulletin 37:185–191
5. Vetter J, Novák P, Wagner MR, Veit C, Möller K-C, Besenhard JO, Winter M,
Wohlfahrt-Mehrens M, Vogler C, Hammouche A (2005) J Power Sources 147:269–
281
6. Wu YP, Rahm E, Holze R (2002) Electrochim Acta 47:3491–3507
7. Liu L, Tian F, Zhou M, Guo H, Wang X (2012) Electrochim Acta 70:360–364
8. Qu Q, Fu L, Zhan X, Samuelis D, Maier J, Li L, Tian S, Li Z, Wu Y (2011) Energy
Environ Sci 4:3985
9. Tang W, Liu LL, Tian S, Li L, Li LL, Yue YB, Bai Y, Wu YP, Zhu K, Holze R (2011)
Electrochemistry Communications 13:1159–1162
10. Ellis BL, Nazar LF (2012) Curr Opin Solid State Mater Sci 16:168–177
11. Chevrier VL, Ceder G (2011) J Electrochem Soc 158:A1011–A1014
12. Ellis BL, Makahnouk WRM, Makimura Y, Toghill K, Nazar LF (2007) Nat Mater
6:749–753
13. Kim S-W, Seo D-H, Ma X, Ceder G, Kang K (2012) Adv Energy Mater 2:710–721
14. Palomares V, Serras P, Villaluenga I, Hueso KB, Carretero-González J, Rojo T (2012)
Energy Environ Sci 5:5884–5901
15. Tarascon J-M, Armand M (2001) Nature 414:359–367
16. Parant J-P, Olazcuaga R, Devalette M, Fouassier C, Hagenmuller P (1971) J Solid
State Chem 3:1–11
17. Delmas C, Fouassier C, Hagenmuller P (1980) Physica B+C 99:81–85
18. Doeff MM, Anapolsky A, Edman L, Richardson TJ, Jonghe LC De (2001) J
Electrochem Soc 148:A230–A236
19. Eriksson TA, Lee YJ, Hollingsworth J, Reimer JA, Cairns EJ, Zhang X, Doeff MM
(2003) Chem Mater 15:4456–4463
[11]
20. Mendiboure A, Delmas C, Hagenmuller P (1985) J Solid State Chem 57:323–331
21. Caballero A, Hernán L, Morales J, Sánchez L, Santos Peña J, Aranda MAG (2002) J
Mater Chem 12:1142–1147
22. Shao-Horn Y (1999) J Electrochem Soc 146:2404–2412
23. Ma X, Chen H, Ceder G (2011) J Electrochem Soc 158:A1307–A1312
24. Kim H, Kim D, Seo D, Yeom M, Kang K, Kim DK, Jung Y (2012) Chem Mater
24:1205–1211
25. Tarascon J, Guyomard D, Wilkens B (1992) Solid State Ionics 57:113–120
26. Sauvage F, Laffont L, Tarascon J-M, Baudrin E (2007) Inorg Chem 46:3289–3294
27. Yang S, Wang X, Wang Y, Chen Q, Li J, Yang X (2010) T Nonferr Metal Soc
20:1892–1898
28. Chick LA, Pederson LR, Maupin GD, Bates JL, Thomas LE, Exarhos GJ (1990) Mater
Lett 10:6–12
29. Berthelot R, Carlier D, Delmas C (2011) Nat Mater 10:74–80
30. Carlier D, Cheng JH, Berthelot R, Guignard M, Yoncheva M, Stoyanova R, Hwang
BJ, Delmas C (2011) Dalton Trans 40:9306–12
31. Dollé M, Hollingsworth J, Richardson TJ, Doeff MM (2004) Solid State Ionics
175:225–228
32. Dollé M, Patoux S, Doeff MM (2005) Chem Mater 17:1036–1043
33. Rietveld HM (1969) J Appl Crystallogr 2:65–71
34. Rietveld HM (1967) Acta Crystallogr 22:151–152
35. Cheary RW, Coelho A (1992) J Appl Crystallogr 25:109–121
36. Lu Z, Donaberger R a., Dahn JR (2000) Chem Mater 12:3583–3590
37. Shu G, Prodi A, Chu S, Lee Y, Sheu H, Chou F (2007) Phys Rev B: Condens Matter
76:184115
38. Meng YS, Hinuma Y, Ceder G (2008) J Chem Phys 128:104708
39. Yabuuchi N, Kajiyama M, Iwatate J, Nishikawa H, Hitomi S, Okuyama R, Usui R,
Yamada Y, Komaba S (2012) Nat Mater 11:512–517
40. Huang Q, Foo ML, Lynn JW, Zandbergen HW, Lawes G, Wang Y, Toby BH, Ramirez
AP, Ong NP, Cava RJ (2004) J Phys Condens Matter 16:5803–5814
[12]
41. Zandbergen H, Foo M, Xu Q, Kumar V, Cava R (2004) Phys Rev B: Condens Matter
70:024101
42. Zhang P, Capaz R, Cohen M, Louie S (2005) Phys Rev B: Condens Matter 71:153102
43. Robertson AD, Armstrong AR, Bruce PG (2001) Chem Mater 13:2380–2386
44. Qu QT, Shi Y, Tian S, Chen YH, Wu YP, Holze R (2009) J Power Sources 194:1222–
1225
45. Wohlfahrt-Mehrens M, Butz A, Oesten R, Arnold G, Hemmer RP, Huggins RA (1997)
J Power Sources 68:582–585
46. Robertson AD, Armstrong a. R, Paterson AJ, Duncan MJ, Bruce PG (2003) J Mater
Chem 13:2367–2373
[13]
List of Figures
Fig. 1 Crystal structure of P2 layered NaxMnO2 and NaxCoyMn1-yO2. Two possible sites for
intercalated Na and three possible ways of filling the octahedral sites (vacancy, Mn, Co) are
shown (see text).
Fig. 2 XRD patterns of (a) NMO and (b) NCO. (c) Rietveld refinement of XRD pattern of
NCO and (d) hkl positions of space group P63mmc.
Fig. 3 FESEM images of NMO (a) and NCO (b); inset in (a) illustrates the layered structure.
Fig. 4 Cyclic voltammograms of a NMO and b NCO (scan rate: 0.3 mV s-1
).
Fig. 5 Galvanostatic charge / discharge profiles of a NMO and b NCO at 0.3 C. 1st, 10
th, and
20th
cycle are shown. Inset: derivative of the charging curves in the range 3.45 V – 3.7 V for
better visibility of features A and B.
Fig. 6 a Integrated charge of galvanostatic discharge at 0.3 C for 20 cycles for both cathode
materials; b Coulomb efficiencies (discharge capacity divided by preceding charge capacity)
for cycles 2…20.
[14]
Fig. 1 Crystal structure of P2 layered NaxMnO2 and NaxCoyMn1-yO2. Two possible sites for
intercalated Na and three possible ways of filling the octahedral sites (vacancy, Mn, Co) are
shown (see text).
Bucher et al., Figure 1
[15]
Fig. 2 XRD patterns of (a) NMO and (b) NCO. (c) Rietveld refinement of XRD pattern of
NCO and (d) hkl positions of space group P63mmc.
Bucher et al., Figure 2
10 20 30 40 50 60 70 80
2 Theta
a
b
c
d
0 0
2
0 0
4
0 1
0
0 1
2
0 1
3
0 0
6
0 1
1
0 1
6
0 1
1
0 0
8
0 1
4
1 1
2
1 1
4
0 2
00
2 1
0 1
8
P2-Na0.7MnO2+z (NMO)
P2-Na0.7Co0.11MnO2+z (NCO)
2
[16]
Fig. 3 FESEM images of NMO (a) and NCO (b); inset in (a) illustrates the layered structure.
Bucher et al., Figure 3
[17]
Fig. 4 Cyclic voltammograms of a NMO and b NCO (scan rate: 0.3 mV s-1
).
Bucher et al., Figure 4
1.5 2 2.5 3 3.5
(b)
0.4
0.3
0.2
0.1
0
-0.1
-0.2
-0.3
(Potential vs. Na|Na+) / V
Cu
rren
t /
mA
P2-Na0.7Co0.11MnO2+z
(NCO)
(a)
0.4
0.2
0.1
0
-0.1
-0.2
-0.3
1.5 2 2.5 3 3.5
0.3
P2-Na0.7MnO2+z
(NMO)
A B
A B
Cycle 1Cycle 2Cycles 3 - 5
0.5
0.5
-0.4
-0.4
Cycle 2
Cycle 3 - 5
Cycle 1
[18]
Fig. 5 Galvanostatic charge / discharge profiles of a NMO and b NCO at 0.3 C. 1st, 10
th, and
20th
cycle are shown. Inset: derivative of the charging curves in the range 3.45 V – 3.7 V for
better visibility of features A and B.
Bucher et al., Figure 5
0 20 40 60 80 100 120 140 160
0 20 40 60 80 100 120 140 160
Capacity / (mA h g-1)
(Po
ten
tial
vs.
Na|
Na+ )
/ V
4.0
3.5
3.0
2.5
2.0
1.5
4.0
3.5
3.0
2.5
2.0
1.5
3.8 V
1.5 V
3.8 V
1.5 V
1st 10st20st
1st10st20st
1st10st20st
1st
(a)
(b)
P2-Na0.7MnO2+z
(NMO)
P2-Na0.7Co0.11MnO2+z
(NCO)
10st20st
AB
1st
3.5 3.6Potential / V
A B
dPo
ten
tial
/ d
Cap
acit
y(a
.u.)
1st
10st
20st
[19]
Fig. 6 a Integrated charge of galvanostatic discharge at 0.3 C for 20 cycles for both cathode
materials; b Coulomb efficiencies (discharge capacity divided by preceding charge capacity)
for cycles 2…20.
Bucher et al., Figure 6
Cycle number
Co
ulo
mb
Eff
icie
ncy
Dis
char
ge C
apac
ity
/ (m
A h
g-1
)
P2-Na0.7MnO2+z
(NMO)
P2-Na0.7Co0.11MnO2+z
(NCO)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20180
160
140
120
100
80
60
40
20
0
100 %
90 %
80 %
70 %
60 %
50 %